Abstract:In this paper, a new feature extraction algorithm is presented, based on a time-frequency representation technique S transform (ST) and two-directional two-dimensional principal component analysis ((2D)(2)PCA), which enables the classification of partial discharge (PD) UHF signals generated by different artificial defects utilizing a support vector machine (SVM) classifier. S transform (ST) is firstly employed to obtain a time-frequency representation of the recorded UHF signals. Then, (2D)(2)PCA is applied to compress the ST amplitude(STA) matrices to extract various feature vectors with different (d(1), d(2)) combinations, i.e. (5, 5), (5, 10), (10, 5) and (10, 10). In the training phase, LIBSVM toolbox is used to construct the SVM classifier, and the two parameters of SVM are selected by particle swarm optimization (PSO) algorithm. After that the PSO-SVM classifier is trained using the selected parameters. During testing, the PD classes are recognized by means of the trained PSO-SVM. Besides, a comparison with the traditional three-layer back propagation neural network (BPNN) is given in this study Experimental results show that the classification accuracies by PSO-SVM are all higher than that by BPNN under four circumstances of (d(1), d(2)) combinations. The success rates of the PSO-SVM with the four feature vectors are above 94% in all cases, and the highest classification accuracy 97.67% is obtained with the (10, 5) combination. Furthermore, by applying (2D)(2)PCA, the feature dimension can be dramatically decreased compared with that of the obtained STA matrices, and the most important information of original data is also preserved. It can he found that the proposed feature extraction and classification algorithm can be effectively applied to PD pattern recognition. Copyright (C) 2012 Praise Worthy Prize S.r.l. - All rights reserved.

Separation and Classification of Concurrent Partial Discharge Signals Using Statistical-Based Feature Analysis

Discrimination of Three or More Partial Discharge Sources by Multi-Step Clustering of Cumulative Energy Features

Separating Multi-Source Partial Discharge Signals Using Linear Prediction Analysis and Isolation Forest Algorithm

Identification of Partial Discharge Sources by Feature Extraction from a Signal Conditioning System

Partial Discharge Signals Separation Using Cumulative Energy Function and Mathematical Morphology Gradient

Waveform Analysis and Feature Extraction of Partial Discharge UHF Signals in Gas Insulated Switchgear

Self-adaptive Separation of Multiple Partial Discharge Sources Based on Optimized Feature Extraction of Cumulative Energy Function

Partial Discharge Diagnostics: Data Cleaning and Feature Extraction

Detection and Identification Techniques for Multi-PD Source in GIS

A hybrid algorithm based on s transform and affinity propagation clustering for separation of two simultaneously artificial partial discharge sources

Classification and separation of partial discharge ultra-high-frequency signals in a 252 kV gas insulated substation by using cumulative energy technique

Characterization of Partial Discharge Signals

Feature Extraction of UHF PD Signals Based on Diode Envelope Detection and Linear Discriminant Analysis

A New Group Of Image Features Derived From Two-Dimensional Linear Discriminant Analysis For Partial Discharge Pattern Recognition

Digital detection, grouping and classification of partial discharge signals at DC voltage

Separation of Multiple Partial Discharge Sources in Power Transformer

Partial Discharge Signal Classification Utilizing Time-Frequency Representation And Two Directional 2dpca

Pulses Separation and Recognition Strategy for Multiple Partial Discharge Sources of Oil-Paper Insulation Based on Time-Frequency Similarity

A Study on the Extraction Method of Partial Discharge Features in Gas Insulated Switchgear Based on Ultra-High Frequency Signal Envelope

Partial Discharge Pattern Recognition Based on 3D Graphs of Phase Resolved Pulse Sequence

Separation of Multiple PD Sources in MV Switchgears during AC Withstand Voltage Test Stage Based on Time-Frequency Representation